A vehicle keyless start matching identification system based on IMMO close-range induction
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XIAMEN AUTOSTAR ELECTRONICS CO LTD
- Filing Date
- 2026-02-10
- Publication Date
- 2026-06-09
Smart Images
Figure CN122166041A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of vehicle electronic anti-theft and short-range communication technology, specifically to a vehicle keyless start matching and identification system based on IMMO near-field sensing. Background Technology
[0002] Vehicle anti-theft systems and keyless start systems typically rely on low-frequency magnetic field induction technology to authenticate the identity between the base station and the electronic key. The vehicle base station drives an LC series resonant circuit through a full-bridge circuit to generate an alternating magnetic field to wake up the electronic key and provide power. The electronic key uses load modulation to feed back its identity data to the base station. This feedback signal manifests at the physical layer as a slight fluctuation in the voltage envelope across the transmitting antenna.
[0003] In existing engineering applications, transmitting antennas need to maintain high current output in the ampere range. The coil wires experience temperature rise due to the Joule heating effect, leading to thermal drift in the DC resistance. This, coupled with voltage fluctuations in the vehicle battery under heavy load, causes the antenna voltage envelope to exhibit a non-linear trend. Current technologies typically employ DC blocking capacitors in the analog front-end or utilize high-pass filters in the digital domain to eliminate baseline drift, and combine this with hardware comparators or real-time edge detection logic to capture instantaneous level transitions to achieve signal synchronization and decoding.
[0004] However, the dynamic changes in coil thermal drift often overlap with the spectrum of the load modulation signal. When processing long bitstream data, the inherent charging and discharging characteristics of high-pass filters easily cause the signal baseline to fluctuate with the data content, leading to signal self-cancellation or phase nonlinear distortion, which in turn causes decoding deviation. Under long-distance or low-power conditions, the modulation depth of the feedback signal is extremely shallow, and the effective amplitude is often submerged in environmental noise. Real-time detection mechanisms that rely on instantaneous levels lack cumulative assessment of waveform energy, and random noise spikes are easily misjudged as valid signal edges, leading to false synchronization of the system and limiting communication distance and robustness. Therefore, this invention provides a vehicle keyless start matching and recognition system based on IMMO proximity sensing to address the shortcomings of existing technologies. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a vehicle keyless start matching and recognition system based on IMMO proximity sensing. This system solves the problems of existing technologies where the charging and discharging characteristics easily cause the signal baseline to fluctuate with the data content, leading to signal self-cancellation or phase nonlinear distortion. Furthermore, under long-distance or low-battery conditions, the modulation depth of the feedback signal is extremely shallow, limiting the communication distance and robustness.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a vehicle keyless start matching and recognition system based on IMMO proximity sensing, comprising: The excitation control module is used to divide the communication interaction process into an instruction sending window, a protocol silence window, and an acknowledgment receiving window according to the communication protocol timing specifications, and to drive the transmitting antenna to establish an induced magnetic field. The signal acquisition module is used to acquire the voltage envelope across the antenna and convert the analog voltage signal into a raw digital sampling sequence. The trend prediction module is used to extract voltage characteristics within the protocol silent window and calculate the reference voltage drift slope parameter that reflects the coil thermal effect and power supply fluctuation. The compensation processing module is used to construct a prediction reference function based on the reference voltage drift slope parameter within the response receiving window, and to perform detrending differential operation on the original digital sampling sequence to output a detrending differential signal; The reconstruction storage module is used to accumulate and reconstruct the detrended differential signal and store it in a circular buffer. The relevant demodulation module is used to perform backtracking matching on the reconstructed sequence in the circular buffer using a preset waveform template to extract the identification code.
[0007] Preferably, the excitation control module includes a protocol timing state machine unit, a waveform generation unit, and a resonance locking unit; The protocol timing state machine unit is used to switch the system state to the protocol silent window at the moment the instruction transmission ends, and to force the waveform generation unit to maintain a constant carrier output state, so as to provide the trend prediction module with a drift learning time base without data modulation. The waveform generation unit is used to construct based on the hardware timer resources of the microcontroller, output two complementary pulse width modulation signals to drive the full-bridge circuit, and insert dead time at the switching edge. The resonant locking unit is used to control the waveform generation unit to perform linear frequency sweep during the communication initialization phase, and to lock the optimal operating frequency based on the current feedback signal of the antenna circuit.
[0008] Preferably, the signal acquisition module includes a precision attenuation network, an envelope detection unit, and a high-speed analog-to-digital conversion unit; The precision attenuation network employs a high-voltage non-inductive resistor divider topology to linearly attenuate the high-voltage resonant signal at both ends of the transmitting antenna. The envelope detection unit is used to demodulate the attenuated high-frequency carrier and extract the voltage envelope signal superimposed with the DC component of coil thermal drift and the AC component of load modulation. The high-speed analog-to-digital conversion unit is used to convert the voltage envelope signal into a raw digital sampling sequence by using the update event of the hardware timer as a trigger source, and transmits the data to the trend prediction module and the compensation processing module through the direct memory access channel.
[0009] Preferably, the trend prediction module includes a sampling window definition unit, a statistical feature extraction unit, and a slope parameter calculation unit; The sampling window delimiting unit is used to define non-overlapping start observation windows and end observation windows within the protocol silent window, and send the oscillation data at the moment of end according to the damping stability margin rejection instruction. The statistical feature extraction unit is used to calculate the average voltage value of the original digital sampling sequence within the starting observation window and the ending observation window, respectively, as the voltage feature value of the starting segment and the voltage feature value of the ending segment. The slope parameter calculation unit is used to calculate the difference between the voltage characteristic value of the ending segment and the voltage characteristic value of the starting segment, and divide it by the time distance between the center points of the two observation windows to obtain the reference voltage drift slope parameter.
[0010] Preferably, the trend prediction module is further configured with parameter validity verification logic; The trend prediction module uses verification logic to determine whether the calculated reference voltage drift slope parameter is within a preset physical reasonable range. The upper limit of the physical reasonableness range is determined based on the maximum temperature rise rate of the coil at the highest ambient temperature and the maximum internal resistance drop rate of the battery, while the lower limit is determined based on the voltage recovery rate at the moment of battery load unloading. When the reference voltage drift slope parameter exceeds the physical reasonableness range, the trend prediction module determines that there is environmental interference and sets the drift slope parameter to zero or maintains the historical value.
[0011] Preferably, the compensation processing module includes a prediction benchmark generation unit and a real-time difference calculation unit; The prediction benchmark generation unit is activated when the response receiving window is entered, and uses the benchmark voltage drift slope parameter and the end segment voltage characteristic value output by the trend prediction module to construct a prediction benchmark function that changes linearly with time. The real-time differential operation unit is used to perform an algebraic subtraction between each raw digital sample value output by the signal acquisition module and the corresponding prediction benchmark function value at the current moment; The real-time differential operation unit is also used to output signed integer data with the sign bit retained as the detrended differential signal, thereby eliminating the DC bias component and linear drift component of the original signal in the digital domain.
[0012] Preferably, the reconstruction storage module includes an accumulation reconstruction unit and a ring buffer management unit; The accumulation and reconstruction unit is used to accumulate multiple continuously input detrended differential signals in segments to generate a high-bit-width reconstruction sequence, wherein the number of accumulated sampling points is set to an integer power of 2. The circular buffer management unit is used to maintain a circular buffer in the random access memory and to continuously write the reconstructed sequence into the buffer using a write pointer; The write pointer is used to automatically update the address after writing data and wrap back to the starting address when it exceeds the buffer boundary, so as to realize the cyclic recording of full-time domain data.
[0013] Preferably, the correlation demodulation module includes a waveform template construction unit and a sliding cross-correlation calculation unit; The waveform template construction unit is used to construct a discretized template sequence with bipolar assignment according to the preamble format and oversampling parameters specified in the communication protocol, and to ensure that the template sequence satisfies the zero mean constraint. The sliding cross-correlation operation unit is used to start after the response receiving window ends and read the reconstruction sequence in the circular buffer; The sliding cross-correlation operation unit is also used to slide the discretized template sequence point by point on the reconstructed sequence and calculate the multiplicative summation value at each sliding position to generate a correlation coefficient sequence.
[0014] Preferably, the related demodulation module further includes an adaptive decision decoding unit, specifically used for: The environmental noise floor is calculated by selecting the data segment corresponding to the end of the protocol silence period in the circular buffer, and a dynamic detection threshold is set based on the noise floor. Search the correlation coefficient sequence for the maximum peak point that exceeds the dynamic detection threshold, and lock the index corresponding to the peak point as the optimal synchronization time; A bit synchronization grid is established based on the optimal synchronization time. Logical numerical decisions are made on subsequent data segments by calculating the integral value of the data in each bit period or the matching degree with the bit symbol template.
[0015] Preferably, the compensation processing module adopts an open-loop deduction architecture; The open-loop extrapolation architecture is used to cut off the feedback path of the received signal to the reference voltage calculation, and only performs unidirectional correction on the signal in the response receiving window based on the physical drift trend learned within the protocol silent window. The real-time differential operation unit is used to maintain the linear trajectory of the prediction reference function unchanged during the generation of the detrended differential signal, regardless of the spectral characteristics of the received signal.
[0016] This invention provides a vehicle keyless start matching and recognition system based on IMMO proximity sensing. It has the following advantages: 1. This invention utilizes the silent window in the communication protocol to extract the voltage drift slope caused by coil thermal effects and power supply fluctuations, and then employs an open-loop prediction method for baseline subtraction in the subsequent receiving window. This mechanism avoids the integral distortion and phase delay caused by the cutoff frequency limitation of traditional high-pass filters for long code stream signals, ensuring that the de-trended differential signal retains the weak load modulation amplitude while maintaining the linearity and flatness of the baseline. This solves the common signal self-cancellation problem in low-frequency inductive communication and guarantees the phase accuracy of subsequent demodulation.
[0017] 2. This invention performs oversampling-based accumulation processing on the detrended signal by reconstructing the storage module. Utilizing the coherence of the effective signal and the incoherence of random noise, the accumulation operation increases the effective bit width of the data while simultaneously increasing the signal amplitude at a ratio superior to the noise increase, thus achieving a signal-to-noise ratio gain in the digital domain. This enables the system to capture weak induced signals with amplitudes lower than the original quantization step, even with limited ADC hardware resolution, reducing the dependence on the gain accuracy of the analog front-end circuit and improving the system's ability to capture weak signals at long distances.
[0018] 3. This invention, by matching and searching the complete reconstructed sequence with a preset waveform template within a circular buffer, focuses the signal energy dispersed in the time domain, thus accurately locking the signal start position even under strong interference conditions with a signal-to-noise ratio below 0dB. This eliminates the impact of transient noise on the false triggering of synchronization logic, improving the anti-interference capability and recognition distance of the vehicle's keyless start matching process. Attached Figure Description
[0019] Figure 1 This is a system architecture diagram of the present invention; Figure 2 This is a flowchart of the method steps of the present invention; Figure 3 The figures are comparison charts of test experimental data for the present invention; wherein, Figure (a) is the original noisy drift signal, Figure (b) is a schematic diagram of the high-pass filter output, and Figure (c) is a schematic diagram of the trend prediction compensation output.
[0020] Among them, 10 is the excitation control module; 20 is the signal acquisition module; 30 is the trend prediction module; 40 is the compensation processing module; 50 is the reconstruction and storage module; and 60 is the correlation demodulation module. Detailed Implementation
[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] See attached document Figure 1 , Figure 1 This is a system architecture diagram according to an embodiment of the present invention. The present invention provides a vehicle keyless start matching and recognition system based on IMMO proximity sensing, which operates inside the vehicle anti-theft controller. The vehicle keyless start matching and recognition system includes an excitation control module 10, a signal acquisition module 20, a trend prediction module 30, a compensation processing module 40, a reconstruction storage module 50, and a correlation demodulation module 60.
[0023] The excitation control module 10 controls the full-bridge drive circuit to generate a low-frequency carrier signal, driving the transmitting antenna to establish an induced magnetic field. According to the communication protocol timing specifications, the excitation control module 10 divides the timeline of a single communication interaction into a command transmission window, a protocol silence window, and a response reception window. Within the command transmission window, the excitation control module 10 controls the carrier signal to modulate and transmit the request command, and within the protocol silence window after the command transmission ends, it maintains a constant amplitude output of the carrier signal.
[0024] The signal acquisition module 20 is connected to the transmitting antenna circuit to acquire the voltage signals across the antenna. The signal acquisition module 20 continuously samples the antenna voltage envelope using an analog-to-digital converter and converts the analog voltage signals into a raw digital sampling sequence. The signal acquisition module 20 is equipped with a direct memory access channel for real-time data transmission.
[0025] The trend prediction module 30 operates within the protocol silent window. It collects voltage characteristic values for the beginning and end segments of the silent window. Based on these collected voltage characteristic values, the trend prediction module 30 calculates the reference voltage drift slope parameter under the current operating condition. The drift slope parameter reflects the linear voltage change trend caused by coil thermal effects and power supply fluctuations.
[0026] The compensation processing module 40 operates within the response receiving window. It receives the drift slope parameter output by the trend prediction module 30 and constructs a prediction benchmark function that changes linearly over time based on the drift slope parameter. The compensation processing module 40 calculates the difference between the original digital sampling sequence and the prediction benchmark function in real time and outputs a detrended differential signal. This process eliminates the DC bias component and linear drift component of the original signal in the digital domain.
[0027] The reconstruction storage module 50 performs virtual oversampling processing on the detrended differential signal. The reconstruction storage module 50 accumulates and reconstructs multiple consecutive detrended differential signal sampling points to generate a high-resolution reconstruction sequence. The reconstruction storage module 50 stores the reconstruction sequence into a circular buffer. The circular buffer is configured in the controller's random access memory.
[0028] The correlation demodulation module 60 operates after the response reception window ends. The correlation demodulation module 60 reads the high-resolution reconstructed sequence from the circular buffer. It then performs a sliding cross-correlation operation on the reconstructed sequence using a preset Manchester waveform template to identify the start position of the valid signal. Based on the identified start position, the correlation demodulation module 60 decodes the data and extracts the identification code.
[0029] See attached document Figure 2 , Figure 2 This is a flowchart of the method steps according to an embodiment of the present invention. The present invention provides a vehicle keyless start matching and identification method based on IMMO proximity sensing, comprising the following steps: S100: Establish a channel and perform a protocol handshake; Excitation control module 10 controls the full-bridge drive circuit to lock the optimal resonant frequency and sends wake-up and challenge commands to the transponder within the command sending window.
[0030] S200, extract drift features; within the protocol silent window after the command is sent, the trend prediction module 30 collects the static features of the coil voltage and calculates the reference voltage drift slope parameter; at this time, the transmitting coil is in full power drive state and there is no data modulation, and the voltage change is mainly caused by physical environmental factors.
[0031] S300, perform blind reception and real-time compensation; within the estimated response reception window, the signal acquisition module 20 starts continuous acquisition; the compensation processing module 40 uses the drift slope parameter calculated in step S200 to eliminate DC bias and linear drift in the original sampled signal in real time and outputs a detrended differential signal.
[0032] S400 performs signal reconstruction and storage; the reconstruction and storage module 50 performs multi-point accumulation reconstruction on the detrended differential signal to increase the effective bit width of the data; the reconstructed data is continuously written to the circular buffer until the response receiving window ends.
[0033] S500 performs backtracking matching demodulation; after data acquisition and storage are completed, the relevant demodulation module 60 performs waveform matching search on the historical data in the circular buffer, extracts the valid identification code and completes identity authentication.
[0034] See attached document Figure 3 The excitation control module 10 drives the subsequent full-bridge drive circuit by generating high-precision logic control signals, thereby exciting the antenna LC resonant circuit to generate a low-frequency magnetic field for communication. In this embodiment, the excitation control module 10 specifically includes a waveform generation unit, a resonance locking unit, and a protocol timing state machine unit.
[0035] The waveform generation unit is built upon the microcontroller's hardware timer resources. To drive the full-bridge circuit, the waveform generation unit outputs two complementary pulse-width modulation (PWM) signals, which are connected to the gate drivers of the left and right half-bridge arms of the full-bridge drive circuit, respectively. Considering the inherent delay in the switching characteristics of the power MOSFETs, to prevent shoot-through of the upper and lower MOSFETs from causing a power supply short circuit, the waveform generation unit inserts a dead time at the switching edges of the two complementary PWM signals. Dead zone time The value is typically set between 0.5 microseconds and 2.0 microseconds, with the specific value depending on the turn-off delay time specified in the datasheet of the selected MOSFET. With descent time The sum is set to reserve sufficient safety margin.
[0036] During the initialization phase before communication establishment, the resonant locking unit is responsible for performing automatic frequency calibration to ensure maximum transmission efficiency. The resonant locking unit controls the waveform generation unit to operate within a preset center frequency range of 125kHz ± 2kHz, using fixed frequency steps of 10kHz or 50kHz. A linear frequency sweep is performed. At each frequency step, the resonant locking unit reads the current feedback signal of the antenna circuit through the analog-to-digital converter interface of the microcontroller. To enhance system robustness and prevent false locking in the event of an antenna open circuit or failure, the resonant locking unit is preset with a minimum current threshold. If all frequency points are scanned across the entire frequency band... All below If the signal is faulty, the system determines that the antenna circuit is faulty and terminates the subsequent process; otherwise, the resonant locking unit uses a peak search algorithm to lock the frequency at [value missing]. Frequency point corresponding to the maximum value The hardware timer's counting period is locked to the count value corresponding to this frequency. For the specific circuit topology of the full-bridge drive circuit and current feedback circuit, those skilled in the art can implement it using conventional H-bridge and sampling resistor schemes, which are well-known technologies in the field and will not be elaborated upon here.
[0037] The protocol timing state machine unit stores timing specifications that conform to the vehicle anti-theft communication protocol. According to the communication protocol, a single complete communication interaction cycle is strictly divided into three continuous time intervals on the time axis, namely the instruction sending window, the protocol silent window, and the response receiving window.
[0038] Upon entering the command transmission window, the protocol timing state machine unit controls the waveform generation unit to modulate the carrier signal. As a preferred implementation, the system employs amplitude shift keying (ASK) or on / off keying (OOK) modulation. Specifically, when transmitting logic "1", the waveform generation unit outputs a PWM signal with a 50% duty cycle, causing the antenna to generate a full-power magnetic field; when transmitting logic "0", the waveform generation unit reduces the PWM duty cycle to a preset modulation depth (e.g., 5%-10%) or stops outputting directly. In this way, the base station sends a wake-up sequence and an encrypted challenge data frame to the electronic key. The duration of the command transmission window... The number of data bits sent and baud rate Decision, that is .
[0039] At the instant the command transmission ends, the protocol timing state machine unit switches the system state to the protocol silent window. This window corresponds to the transponder's internal processing time or capacitor charging time specified in the communication protocol. During this period, the protocol timing state machine unit forces the waveform generation unit to maintain a constant output state, i.e., a continuous output with a fixed duty cycle (e.g., 50%) and a frequency locked. The unmodulated carrier signal. At this time, the transmitting antenna is in a continuous high-current operating state, according to Joule's law. This leads to an increase in coil temperature and DC resistance. Meanwhile, the vehicle power supply experienced a voltage drop due to heavy load. Since there is no data modulation during this window, the change in the antenna voltage envelope purely reflects the physical characteristics of thermal drift and power fluctuations, providing an ideal learning time base for subsequent modules. The duration of this window... Typically, the protocol standard presets a fixed value, such as 2 to 5 milliseconds, which is sufficient for the system to capture an effective voltage drift slope.
[0040] After the protocol silence window ends, the protocol timing state machine unit control system enters the response reception window. Within this window, the waveform generation unit continues to transmit an unmodulated constant carrier to provide the energy carrier required for load modulation of the electronic key via magnetic field coupling. At this time, the excitation control module 10 maintains the steady state of the physical layer, providing a background reference for the signal acquisition module 20 to extract the weak load modulation signal. The duration of the response reception window... Based on the longest possible response time setting of the electronic key, it is usually set to... This is to ensure that reception is not interrupted.
[0041] The signal acquisition module 20 connects the hardware output of the excitation control module 10 to the digital processing core of the vehicle controller, and consists of a high dynamic range analog front-end circuit and a high-speed digital conversion interface. Since the keyless start system in this embodiment uses a series resonance method to drive the transmitting antenna to obtain the maximum magnetic field strength, the resonant voltage amplitude across the antenna typically reaches several hundred volts (e.g., 300Vpp) when operating at full power. Given that this voltage amplitude far exceeds the tolerance limit of conventional microcontroller I / O ports (typically 3.3V or 5V), the signal acquisition module 20 is sequentially configured with a precision attenuation network, an envelope detection unit, and a high-speed analog-to-digital conversion unit along the hardware signal path.
[0042] To safely introduce high-voltage signals into the processing system, the precision attenuation network employs a high-voltage non-inductive resistive voltage divider topology. As a preferred implementation, the voltage divider resistors are selected as non-inductive resistors with low temperature coefficients and minimal parasitic inductance to prevent additional phase delay or resonance introduced by parasitic inductance under high-frequency carrier waves, ensuring that the attenuated signal phase remains strictly synchronized with the antenna current.
[0043] The linearly attenuated signal is fed into an envelope detection unit, which is configured as a Schottky diode with low forward voltage drop or a precision detection circuit composed of a high-bandwidth operational amplifier. The physical purpose of this detection unit is to demodulate the 125kHz high-frequency carrier and extract the voltage envelope signal reflecting changes in magnetic field energy. This voltage envelope signal superimposed in the time domain contains two key types of information: first, a DC component corresponding to the coil's steady-state current and ambient thermal drift, which has a large amplitude and changes slowly; second, a weak AC component corresponding to the electronic key load modulation action, whose amplitude is usually submerged in the DC component and noise.
[0044] After analog signal conditioning, the high-speed analog-to-digital converter (ADC) uses the successive approximation (SAR) analog-to-digital converter (ADC) integrated within the microcontroller to convert the voltage envelope into a digital sequence. To accommodate the virtual oversampling algorithm used in the subsequent signal reconstruction and buffering module 50, the ADC configuration strategy in this embodiment breaks through the minimum requirements of the conventional Nyquist sampling theorem and instead adopts a high-rate oversampling strategy. Regarding the sampling triggering mechanism, the system strictly prohibits software polling triggering; instead, the ADC conversion trigger source is hardware-bound to the update event of a high-precision timer. The significant advantage of this hardware triggering mechanism is that it can control the sampling aperture jitter within the nanosecond range, ensuring the sampling time interval... This ensures strict consistency, thereby avoiding nonlinear calculation errors introduced in subsequent linear trend prediction calculations due to time base jitter.
[0045] For ADC sampling frequency The specific settings must be based on a strict mathematical relationship between the communication baud rate and the requirements of subsequent digital signal processing. Specifically, the sampling rate needs to take into account the baud rate specified in the communication protocol. (e.g., 4kbps Manchester encoding), the accumulation factor set by the signal reconstruction module. And the number of valid samples required for each data bit after reconstruction. To ensure that after passing through After multiple rounds of downsampling, each data bit still retains a sufficient number of discrete points. To describe waveform details and support phase synchronization, sampling frequency Determined based on the following inequality: ; In the formula: The data transmission rate specified by the communication protocol; In this embodiment, the accumulation factor set in the subsequent reconstruction storage module 50 is used as the basis for... The preferred value range is an integer between 16 and 64. The physical meaning is that by accumulating the values, the random quantization noise is averaged out in exchange for... Additional resolution gain of bits; To ensure that the demodulator can accurately capture signal transition edges, the minimum number of samples contained in each bit after reconstruction is typically set to [a certain value]. Based on the above parameter logic, if the system is set... , and If so, the physical sampling rate of the ADC needs to be set to 512kHz or higher.
[0046] This high-frequency sampling generates a large data stream, which would exhaust computing power if directly handled by the central processing unit. Therefore, the signal acquisition module 20 is equipped with a Direct Memory Access (DMA) controller. The DMA controller is configured in circular buffer mode, meaning that when the pre-allocated memory buffer is full, the DMA pointer automatically wraps back to the beginning address of the buffer to continue writing. This mechanism ensures that the original digital sampling sequence remains intact throughout the entire communication process, covering command sending, protocol silence, and acknowledgment reception. It can be recorded continuously and without omission, providing a complete time-domain data foundation without time breakpoints for the "backtracking" processing of subsequent modules.
[0047] The trend prediction module 30 is used to extract and mathematically model environmental physical features within the protocol silence window. In this embodiment, the trend prediction module 30 is based on the physical characteristics of low-frequency inductive communication: when the excitation control module 10 ends command transmission and enters the silence period, although the logic control signal remains constant, the actual voltage envelope of the antenna circuit... It is not an ideal DC level, but rather exhibits a dynamic characteristic that changes slowly over time. This dynamic change is mainly controlled by two physical processes: First, when the transmitting coil is driven by a large current of 2A-5A, the temperature of its copper wire increases with the accumulation of Joule heat. According to the temperature coefficient of resistance of copper (approximately 0.00393 / ℃), this leads to an increase in the equivalent series resistance (ESR), which in turn causes a linear decrease in the circuit current and induced voltage. Secondly, during continuous heavy-load discharge, the terminal voltage of the vehicle battery exhibits a slight drop trend due to the polarization effect of the battery's internal resistance.
[0048] To quantify the aforementioned physical drift in the digital domain, the trend prediction module 30 employs a first-order linear regression model based on short-window statistical analysis. Logically, the trend prediction module 30 comprises three processing units: sampling window definition, statistical feature extraction, and slope parameter calculation.
[0049] The sampling window definition unit defines the sampling window based on the protocol timing, within the protocol silent window. The system internally defines two non-overlapping observation sub-intervals. To ensure the purity of the sampled data, the system needs to eliminate the influence of transient responses on the time axis; specifically... At the instant the timing command ends, the LC resonant circuit will generate damped oscillations. To avoid this nonlinear oscillation region, the opening point of the initial observation window is set to... Here, the damping stability margin The value is determined based on the quality factor of the antenna circuit. and carrier frequency Determined, must meet the following requirements (i.e., more than 3 to 5 oscillation decay time constants) This value is typically set to 20 to 50 sampling points. Similarly, the end point of the observation window is set to... To reserve a safety guardrail to prevent data aliasing caused by the premature arrival of the next stage response signal.
[0050] The statistical feature extraction unit is responsible for calculating the voltage statistical characteristics within these two observation windows. Given that single-point sampled data is highly susceptible to random interference from Gaussian white noise and quantization noise, this unit abandons the single-point difference method and instead uses the interval averaging method to extract the central trend of the reference voltage, thereby improving the signal-to-noise ratio. Specifically, the average voltage characteristic value of the initial observation window... The calculation is as follows: ; Accordingly, the average voltage characteristic value at the end of the observation window The calculation is as follows: ; In the formula: The original digital sampling sequence output by the signal acquisition module 20 has a value range that depends on the bit width of the ADC (e.g., 0 to 4095). The observation window length is preferably set to [value]. (e.g., 32 or 64) so that the microcontroller can replace division operations with shift instructions, reducing the computation cycle consumption.
[0051] Based on the two noise-smoothed steady-state characteristic points mentioned above, the slope parameter calculation unit constructs a voltage drift model under the current operating condition. This voltage drift model is based on the short-time linear approximation principle, that is, within the protocol silence period of tens of milliseconds, the second-order exponential term of thermal effects is ignored, and the total drift is regarded as a linear function of time. The slope parameter calculation unit calculates the drift slope of the reference voltage relative to the sampling time index. The calculation formula is as follows: ; In the formula: The time distance between the center points of the two observation windows, i.e. To ensure the numerical stability of the algorithm, the system needs to perform a certain operation before executing the division. Perform a non-zero check to ensure that the protocol's silent window length is sufficient to accommodate two observation windows and the guardrail (i.e.) Otherwise, a timing configuration error will be marked.
[0052] The drift slope was calculated. Subsequently, the trend prediction module 30 further performs parameter validity verification. The system has a preset physical reasonableness range. This threshold range is calculated based on the system's extreme operating conditions: upper limit This corresponds to the sum of the maximum temperature rise rate of the coil when it operates at full power under the highest ambient temperature and the maximum internal resistance drop rate of the battery. lower limit Typically, a negative value is used, corresponding to the possible voltage rebound during the moment the battery load is unloaded. If the calculated... Falling outside this range (e.g., due to strong external electromagnetic pulse interference) (Abnormal mutation), the system determines the current The value is invalid, and it will be forced to be set to zero or the historical check value from the previous period will be used.
[0053] This protection logic prevents erroneous drift compensation parameters from introducing artificial reference bias in subsequent stages, thereby ensuring the system's robustness in harsh electromagnetic environments. The final determined... Value and characteristic value of the end segment It is latched into the shared memory area and used as input parameters for the subsequent dynamic baseline compensation module.
[0054] The compensation processing module 40 is used to perform open-loop signal compensation based on a prediction model within the response reception window. In this embodiment, the design of the compensation processing module 40 is based on the principle of physical inertia in low-frequency induction systems: that is, within a short time window of milliseconds, thermal drift caused by coil temperature and voltage drop in battery are continuous and do not undergo abrupt changes. In the response reception stage, the system abandons the traditional closed-loop feedback filtering architecture and instead adopts a parameter-locked open-loop extrapolation architecture. This open-loop extrapolation architecture aims to solve the baseline drift or signal self-cancellation phenomenon caused by the cutoff frequency limitation of traditional high-pass filters when the received Manchester encoded signal contains long strings of "0"s or "1"s (e.g., a synchronization header or a specific ID segment).
[0055] The compensation processing module 40 includes a prediction reference generation unit and a real-time differential operation unit, which work together to achieve signal detrending processing.
[0056] The prediction baseline generation unit enters the response reception window (i.e., time index) in the system timing. The system starts running at [time]. The prediction benchmark generation unit reads the drift slope parameters latched by the trend prediction module 30. and the average voltage characteristic value at the end of the observation window Based on the physical assumption that the current benchmark is equal to the sum of the previous steady-state value and the accumulated drift, the prediction benchmark generation unit constructs a prediction benchmark function that dynamically changes over time. Its mathematical calculation model is as follows: ; In the formula: This is the index value of the current sampling point on the global time axis, and its value range covers the entire response reception window; For calculation The reference time anchor point, in this embodiment, is defined as the center position index of the end observation window, i.e. The product term in this formula This represents the linear drift increment from the observation time to the current time. To prevent computational overflow during the processing of long data frames (such as those exceeding 100ms), the arithmetic unit typically performs this multiply-accumulate operation in a 32-bit or higher register.
[0057] The real-time differential processing unit keeps synchronized with the data stream of the signal acquisition module 20. Whenever the analog-to-digital converter (ADC) outputs a new raw digital sample value... The real-time differential operation unit immediately performs algebraic difference operations to output a detrended differential signal. The formula for difference operations is defined as follows: ; In the formula: This is the output difference sequence. Due to the original signal... Possibly less than the forecast baseline (For example, during the negative half-cycle of the signal or under noise disturbance). It must be defined as a signed integer format, such as 16-bit or 32-bit two's complement, to fully preserve the polarity information of the signal.
[0058] Through the above processing, the output is... Numerically, the common-mode DC component introduced by the hundreds of volt carrier wave and the linear trend term caused by environmental factors are eliminated, retaining only the weak amplitude fluctuations caused by electronic key load modulation and high-frequency noise. Even long pulse signals with extremely low frequencies, as long as their amplitude deviates from the predicted physical baseline, will be completely mapped into the differential result without waveform distortion caused by filter phase nonlinearity. This characteristic ensures the phase accuracy of subsequent demodulation modules when processing weak signal edges.
[0059] The reconstruction storage module 50 connects the high-speed signal processing unit at the front end and the low-speed decoding unit at the back end. In this embodiment, the reconstruction storage module 50 does not directly transmit the high-speed raw differential stream output by the compensation processing module 40. Instead, based on the principles of oversampling and decimation, it converts the high-frequency, low-bit-width sampled stream into a low-frequency, high-bit-width reconstruction sequence. The physical basis of this processing lies in the incoherence of Gaussian white noise and the coherence of the effective signal: by accumulating the values of multiple consecutive sampling points, the amplitude of the effective signal increases linearly, while the amplitude of the random noise superimposed on it increases in a square root manner, thereby achieving a signal-to-noise ratio (SNR) gain in the digital domain that is difficult for analog filters to achieve. The reconstruction storage module 50 includes an accumulation reconstruction unit and a ring buffer management unit.
[0060] The data output clock of the accumulation and reconstruction unit is synchronized with that of the real-time differential operation unit. This takes into account the input differential signal. Since the data is signed and the accumulation process can lead to numerical inflation, the accumulation reconstruction unit is internally configured with a signed accumulation register with a width of at least 32 bits to prevent data overflow. In specific implementation, the accumulation reconstruction unit uses a preset accumulation factor... For continuously input detrended differential signals Perform segmented accumulation to generate a reconstructed sequence. Its mathematical operation model is as follows: ; In the formula: The index for reconstructing the sequence is updated at a frequency equal to the original sampling rate. of ; The input is a detrended differential signal sequence; The accumulation factor (or extraction rate) is strictly limited to integer powers of 2 (e.g., 16, 32, 64) in this embodiment. Integer powers of 2 are chosen as... The engineering basis for this is that microprocessors can use single-cycle hardware right shift instructions to replace time-consuming division operations for normalization, while also facilitating counter toggling control through bitmask operations. When At this time, merging every 16 original sampling points into one reconstruction point can increase the effective data bit width. This allows the system to distinguish weak signal changes with amplitudes smaller than the original quantization step of the ADC.
[0061] The generated reconstructed sequence after the accumulation operation is completed The data is transmitted to the circular buffer management unit. Because the length of the response data of the vehicle keyless start system is uncertain, and the subsequent demodulation algorithm needs to perform backtracking analysis after receiving the complete data packet, the system allocates a contiguous memory space in the microcontroller's random access memory as a circular buffer.
[0062] The circular buffer management unit maintains a write pointer. This pointer always points to the currently writable memory address in the buffer. The update logic for the write pointer is as follows: ; In the formula: The total depth of the circular buffer (in words). This indicates the current write pointer position. To ensure sufficiency of disclosure, this embodiment provides... Minimum computational constraints: .in, The maximum physical frame length defined for the communication protocol (e.g., 150ms). This represents the effective data rate after reconstruction. If the reconstructed sampling rate is 32kHz (i.e., each Manchester symbol contains 8 sampling points) and the maximum frame length is 150ms, then... It needs to be configured to hold at least 4800 words. The modulo operation in the formula ensures that when the amount of data written exceeds the physical boundary of the buffer, the pointer can automatically wrap back to the starting address, achieving cyclic overwriting of the oldest historical data.
[0063] The correlation demodulation module 60 is configured as the back-end processing unit of the physical layer of the vehicle keyless start system. The operation mechanism of this correlation demodulation module 60 adopts a batch processing mode based on store-backtracking. In this embodiment, the system does not perform real-time streaming demodulation while receiving data; instead, it waits for the protocol timing state machine unit to indicate the end of the response reception window, and for the ring buffer management unit to have locked the complete reconstruction sequence. Afterwards, the relevant demodulation modules 60 were started and running.
[0064] In long-distance scenarios of low-frequency inductive communication, the effective signal is often submerged in background noise of comparable or even higher amplitude (i.e., SNR < 0dB). Traditional edge detection techniques based on instantaneous level transitions are ineffective. It is necessary to utilize the energy accumulation characteristics of the known signal waveform on the time axis and achieve signal energy focusing and extraction through cross-correlation operations. The correlation demodulation module 60 includes a waveform template construction unit, a sliding cross-correlation operation unit, and an adaptive decision decoding unit.
[0065] The waveform template construction unit is responsible for generating a standard digital reference sequence for matching. This unit constructs a discretized template sequence based on the preamble format specified by the communication protocol (e.g., continuous high levels or a specific synchronization sequence) and the current oversampling parameters of the system. To maximize the dynamic range of cross-correlation operations under DC conditions, the template sequence in this embodiment... A bipolar assignment strategy is adopted, instead of the traditional unipolar (0 / 1) assignment. Specifically, the sampling point corresponding to logic "1" is assigned a value... The sampling point corresponding to logic "0" is assigned a value of -1, and it is ensured that To eliminate the effects of DC bias. Assume the preamble contains... There are bits, and the number of reconstruction sampling points for each bit is . (Based on sampling rate) baud rate and cumulative factor Decision, that is If the template length is set to ), then the template length is set to . .
[0066] Based on the constructed template, the sliding cross-correlation unit performs a full-time domain matching search. This unit will use the local template... In the reconstructed sequence The algorithm slides point by point upwards, calculating the correlation energy value at each time offset. The mathematical model for cross-correlation is as follows: ; In the formula: The starting index of the sliding window in the reconstruction sequence is [index], and its traversal range is [range]. ; Indicates the reconstructed sequence in The sampled value at the location; This is the output correlation coefficient sequence. The formula measures the phase consistency between the received signal segment and the standard waveform. When the signal waveform in the reconstructed sequence is perfectly phase-aligned with the template waveform, all positive and negative peaks correspond one-to-one, the product term is always positive, and after accumulation... It exhibits an energy peak; conversely, for uncorrelated noise segments, the product terms are randomly distributed and cancel each other out, resulting in... It remains at a low level.
[0067] In obtaining the correlation coefficient sequence Subsequently, the adaptive decision decoding unit performs signal acquisition. To adapt to the differences in electromagnetic noise floor under different vehicle environments, this unit employs dynamic threshold determination logic.
[0068] First, the adaptive decision decoding unit estimates the noise floor of the current environment. In a preferred embodiment, the system selects the first data segment of the circular buffer (corresponding to the end of the protocol silence period before a response has been sent) as the noise observation window, and calculates the average absolute amplitude of the data within this window: ; In the formula: The length of the noise observation window is typically between 128 and 256 sampling points. The first in the data sequence The value of each point, This represents the summation index variable. Based on this noise floor, the system sets a dynamic detection threshold. .in, This is the signal-to-noise ratio sensitivity coefficient, and its value is typically set between 3.0 and 6.0. The determination of this coefficient is based on the false alarm rate constraint under Gaussian noise background. The higher the value, the lower the probability of a false alarm, but the ability to detect weak signals decreases accordingly.
[0069] Adaptive decision decoding unit in Search in the middle satisfies The maximum peak point is used as the index to lock the optimal synchronization time. After locking the synchronization point, the system establishes a bit synchronization grid to... Using sampling points as the step size, integral decision decoding is performed on subsequent data load segments. The decision formula for each data bit is as follows: ; In the formula: For the first Energy integral value per bit period For data bit index, The sampling step size, For the variable to be summed during integration, For the optimal synchronization time, Indicates the template length. If... If so, then the logical value of that bit is determined to be "1"; if If the condition is not met, the decision is "0" (assuming a specific phase of non-return-to-zero code or Manchester code is used). This integral decision method utilizes the energy of the entire symbol period, which, compared to single-point sampling decision, can effectively smooth high-frequency glitches within the symbol period, thereby reducing the bit error rate (BER). The final parsed binary ID data stream is transmitted to the application layer of the vehicle controller for encryption verification and authentication.
[0070] To better understand the technical solution of this invention, the following description is based on a specific application scenario.
[0071] Application scenario configuration: Vehicle environment: The target vehicle is equipped with a 12V power system, and the inductance of the transmitting antenna is 345. Quality factor .
[0072] Controller hardware: It adopts a 32-bit automotive-grade microcontroller with a main frequency of 80MHz and a built-in 12-bit SARADC.
[0073] Communication protocol: Manchester encoding, baud rate (i.e., bit width 250) ).
[0074] System parameters: Accumulation factor Number of reconstructed samples for each digit ADC sampling rate The protocol's silent window duration is set to 3ms.
[0075] Implementation process: Quiet period feature extraction: After the controller finishes sending the wake-up command, it enters a 3ms protocol quiet window. During this time, due to the high current heating of the coil, the voltage envelope at both ends of the antenna linearly decreases from 300.5V to 300.2V within 3ms.
[0076] Trend Modeling: The trend prediction module 30 collects the average voltage at 0.5ms of the silent window (the initial observation window). Average voltage was collected at 2.5ms. The drift slope was calculated. .
[0077] Open-loop compensation: A 100ms response reception window then begins. At this time, although there is no new reference, the compensation processing module 40 operates based on the locked... At 50ms, automatically subtract from the reference voltage The predicted drift amount.
[0078] Although the background voltage of the original signal at 50ms has drifted by 7.5V due to thermal effects (far greater than the effective signal amplitude of only 0.5V), the drift is completely canceled out after subtracting the prediction reference. The weak 0.5V effective signal is completely preserved in the differential sequence, ensuring the accuracy of subsequent demodulation.
[0079] Experimental verification: Experimental environment setup: Signal source: Simulated 125kHz carrier envelope, superimposed with Manchester-coded ID data.
[0080] Interference conditions: Strong linear drift: rapid heating of the analog coil, superimposed with a DC baseline drift with a slope of -20V / s; Extremely low signal-to-noise ratio: Gaussian white noise is added, and the signal-to-noise ratio is set. .
[0081] Comparison objects: Existing technology: A first-order IIR digital high-pass filter with a cutoff frequency of 100Hz is used for DC removal; This invention employs the trend prediction and open-loop compensation method described in this embodiment.
[0082] See attached document Figure 3 The figure shows a comparison of the results of the two solutions under the same severe working conditions. According to... Figure 3 (a) It can be seen that the original signal exhibits a downward sloping trend, and the effective waveform is covered by noise, making it difficult to distinguish data bits with the naked eye. According to Figure 3 (b) After being processed by the high-pass filter, some DC is eliminated, but there are two problems: one is phase nonlinearity, which leads to blurred waveform edges; the other is baseline jitter. When the data bit stream has continuous phase changes, the filter tries to track the signal itself, causing the output waveform to fluctuate up and down. The integral energy of some data bits is incorrectly canceled, ultimately resulting in a demodulation bit error rate as high as 15%. Figure 3 (c) After applying open-loop compensation, the baseline of the output differential signal is absolutely flat. Since the drift path is accurately predicted by using a silent window, the signal after subtracting the reference retains only the pure modulation transition and high-frequency noise. The waveform phase is strictly aligned with the original data to verify that the signal restoration fidelity of this scheme is better than that of traditional technology under extreme conditions where the signal-to-noise ratio is less than 0dB and accompanied by thermal drift.
[0083] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A vehicle keyless start matching and recognition system based on IMMO proximity sensing, characterized in that, include: The excitation control module is used to divide the communication interaction process into an instruction sending window, a protocol silence window, and an acknowledgment receiving window according to the communication protocol timing specifications, and to drive the transmitting antenna to establish an induced magnetic field. The signal acquisition module is used to acquire the voltage envelope across the antenna and convert the analog voltage signal into a raw digital sampling sequence. The trend prediction module is used to extract voltage characteristics within the protocol silent window and calculate the reference voltage drift slope parameter that reflects the coil thermal effect and power supply fluctuation. The compensation processing module is used to construct a prediction reference function based on the reference voltage drift slope parameter within the response receiving window, and to perform detrending differential operation on the original digital sampling sequence to output a detrending differential signal; The reconstruction storage module is used to accumulate and reconstruct the detrended differential signal and store it in a circular buffer. The relevant demodulation module is used to perform backtracking matching on the reconstructed sequence in the circular buffer using a preset waveform template to extract the identification code.
2. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 1, characterized in that, The excitation control module includes a protocol timing state machine unit, a waveform generation unit, and a resonance locking unit; The protocol timing state machine unit is used to switch the system state to the protocol silent window at the moment the instruction transmission ends, and to force the waveform generation unit to maintain a constant carrier output state, so as to provide the trend prediction module with a drift learning time base without data modulation. The waveform generation unit is used to construct based on the hardware timer resources of the microcontroller, output two complementary pulse width modulation signals to drive the full-bridge circuit, and insert dead time at the switching edge. The resonant locking unit is used to control the waveform generation unit to perform linear frequency sweep during the communication initialization phase, and to lock the optimal operating frequency based on the current feedback signal of the antenna circuit.
3. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 1, characterized in that, The signal acquisition module includes a precision attenuation network, an envelope detection unit, and a high-speed analog-to-digital conversion unit. The precision attenuation network employs a high-voltage non-inductive resistor divider topology to linearly attenuate the high-voltage resonant signal at both ends of the transmitting antenna. The envelope detection unit is used to demodulate the attenuated high-frequency carrier and extract the voltage envelope signal superimposed with the DC component of coil thermal drift and the AC component of load modulation. The high-speed analog-to-digital conversion unit is used to convert the voltage envelope signal into a raw digital sampling sequence by using the update event of the hardware timer as a trigger source, and transmits the data to the trend prediction module and the compensation processing module through the direct memory access channel.
4. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 1, characterized in that, The trend prediction module includes a sampling window definition unit, a statistical feature extraction unit, and a slope parameter calculation unit. The sampling window delimiting unit is used to define non-overlapping start observation windows and end observation windows within the protocol silent window, and send the oscillation data at the moment of end according to the damping stability margin rejection instruction. The statistical feature extraction unit is used to calculate the average voltage value of the original digital sampling sequence within the starting observation window and the ending observation window, respectively, as the voltage feature value of the starting segment and the voltage feature value of the ending segment. The slope parameter calculation unit is used to calculate the difference between the voltage characteristic value of the ending segment and the voltage characteristic value of the starting segment, and divide it by the time distance between the center points of the two observation windows to obtain the reference voltage drift slope parameter.
5. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 4, characterized in that, The trend prediction module is also configured with parameter validity verification logic. The trend prediction module uses verification logic to determine whether the calculated reference voltage drift slope parameter is within a preset physical reasonable range. The upper limit of the physical reasonableness range is determined based on the maximum temperature rise rate of the coil at the highest ambient temperature and the maximum internal resistance drop rate of the battery, while the lower limit is determined based on the voltage recovery rate at the moment of battery load unloading. When the reference voltage drift slope parameter exceeds the physical reasonableness range, the trend prediction module determines that there is environmental interference and sets the drift slope parameter to zero or maintains the historical value.
6. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 1, characterized in that, The compensation processing module includes a prediction benchmark generation unit and a real-time difference calculation unit; The prediction benchmark generation unit is activated when the response receiving window is entered, and uses the benchmark voltage drift slope parameter and the end segment voltage characteristic value output by the trend prediction module to construct a prediction benchmark function that changes linearly with time. The real-time differential operation unit is used to perform an algebraic subtraction between each raw digital sample value output by the signal acquisition module and the corresponding prediction benchmark function value at the current moment; The real-time differential operation unit is also used to output signed integer data with the sign bit retained as the detrended differential signal, thereby eliminating the DC bias component and linear drift component of the original signal in the digital domain.
7. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 1, characterized in that, The reconstructed storage module includes an accumulation reconstructed unit and a ring buffer management unit; The accumulation and reconstruction unit is used to accumulate multiple continuously input detrended differential signals in segments to generate a high-bit-width reconstruction sequence, wherein the number of accumulated sampling points is set to an integer power of 2. The circular buffer management unit is used to maintain a circular buffer in the random access memory and to continuously write the reconstructed sequence into the buffer using a write pointer; The write pointer is used to automatically update the address after writing data and wrap back to the starting address when it exceeds the buffer boundary, so as to realize the cyclic recording of full-time domain data.
8. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 1, characterized in that, The related demodulation module includes a waveform template construction unit and a sliding cross-correlation calculation unit; The waveform template construction unit is used to construct a discretized template sequence with bipolar assignment according to the preamble format and oversampling parameters specified in the communication protocol, and to ensure that the template sequence satisfies the zero mean constraint. The sliding cross-correlation operation unit is used to start after the response receiving window ends and read the reconstruction sequence in the circular buffer; The sliding cross-correlation operation unit is also used to slide the discretized template sequence point by point on the reconstructed sequence and calculate the multiplicative summation value at each sliding position to generate a correlation coefficient sequence.
9. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 8, characterized in that, The related demodulation module further includes an adaptive decision decoding unit, specifically used for: The environmental noise floor is calculated by selecting the data segment corresponding to the end of the protocol silence period in the circular buffer, and a dynamic detection threshold is set based on the noise floor. Search the correlation coefficient sequence for the maximum peak point that exceeds the dynamic detection threshold, and lock the index corresponding to the peak point as the optimal synchronization time; A bit synchronization grid is established based on the optimal synchronization time. Logical numerical decisions are made on subsequent data segments by calculating the integral value of the data in each bit period or the matching degree with the bit symbol template.
10. The vehicle keyless start matching and recognition system based on IMMO proximity sensing according to claim 6, characterized in that, The compensation processing module adopts an open-loop deduction architecture; The open-loop extrapolation architecture is used to cut off the feedback path of the received signal to the reference voltage calculation, and only performs unidirectional correction on the signal in the response receiving window based on the physical drift trend learned within the protocol silent window. The real-time differential operation unit is used to maintain the linear trajectory of the prediction reference function unchanged during the generation of the detrended differential signal, regardless of the spectral characteristics of the received signal.